Validation of the I- and D-type epistemic curiosity scale among young Chinese children and implications on early curiosity nurture

Curiosity is a universal motivation to seek novel information and close knowledge gaps and is linked to personality, learning, and well-being. Epistemic curiosity, elicited by information gaps, comprises Interest (I; joyous exploration for learning novelty) and Deprivation (D; drive to resolve uncertainty and alleviate aversive feelings) subtypes. Adult studies show distinct associations for I and D with personality traits, emotions, academic achievement, and creativity. However, empirical evidence for epistemic curiosity in early childhood, especially in non-Western contexts, is limited, and measures developed in Western cultures may not generalize to collectivist contexts like China. The study aims to translate and validate the I/D-YC scale in Chinese preschoolers, test its factor structure, reliability, measurement invariance, and construct validity, and examine associations with temperament, inhibitory control, supportive parenting, and creativity within the Chinese cultural context.

Curiosity has been linked with openness, extraversion, ambiguity tolerance, creativity, positive emotions, broader knowledge, and better memory retention. Epistemic curiosity entails higher-order information seeking shared by humans and some animals and relates to the development of knowledge networks. Adult research distinguishes I-type (joyous exploration) from D-type (deprivation sensitivity), with I associated with openness and happiness and D with conscientiousness and anxiety; I relates directly to academic performance and creative problem-solving, whereas D shows indirect or narrower associations. Child curiosity is influenced by cultural norms; traditional societies may encourage indirect information seeking, with prior evidence mainly from Western samples. The original I/D-YC (10 items) showed good internal consistency in Dutch samples but failed to differentiate I and D in Swedish samples. Several curiosity measures have shown structural differences in Chinese populations (e.g., CEI-II collapsing to one factor). There is a need for culturally sensitive instruments for Chinese preschoolers to capture epistemic curiosity reliably.

Translation: Two PhD students and a vice-professor in psychology translated the I/D-YC from English to Chinese; an English teacher back-translated to ensure semantic equivalence. The final Chinese scale retained 10 items across two factors (I and D), consistent with the original. Participants: Data were collected online due to COVID-19 restrictions across four samples. Sample 1: 111 parents (children mean age 4.80, SD=0.97) from a Haikou kindergarten; 2 responses removed for short times; used for EFA. Sample 2: 389 parents from kindergartens in Chongqing, Sichuan, and Hainan (children mean age 4.30, SD=0.90; 52.70% boys); used for CFA, reliability, and 2-week test-retest with 82 parents (children mean age 3.49, SD=0.40; 58.54% boys). Power analysis (WebPower) suggested minimum n=305 for CFA. Sample 3: 189 parents (children mean age 3.90, SD=1.25; 53.44% boys) after excluding 14 for polygraph errors; used for construct validity. Sample 4: 129 teachers (children mean age 4.92, SD=0.69; 53.49% boys); used for convergent validity via teacher report. Measures: Chinese I/D-YC (10 items; 4-point Likert 1–4; higher scores=greater epistemic curiosity) in Samples 1–3. CBQ short subscales (6 items each; 7-point Likert) for shyness (α=0.81), high-intensity pleasure (α=0.76), perceptual sensitivity (α=0.73) in Sample 3. Parent Temperament Questionnaire—Approach/Withdrawal (8 items; 7-point Likert; α=0.63) in Sample 3. CHEXI inhibition and regulation subscales (6 and 5 items; 5-point Likert; higher scores=greater deficits; α=0.84) in Sample 3. Brief Williams Creativity Scale (4 items; 3-point Likert; α=0.80) in Sample 3. Teachers Questionnaire on Young Child’s Curiosity (brief 8-item teacher-rated; 4-point Likert; α=0.92) in Sample 4. Demographics: SES (7 items; α=0.61), Supportive Parenting (4 items; α=0.85) in Sample 3. Statistical analysis: Conducted in SPSS 25.0 and R 4.3.2 (psych, MVN, lavaan). EFA (Sample 1): normality checks showed univariate normality but not multivariate; used principal-axis factoring with oblique Oblimin rotation. Factor number determined via parallel analysis and Velicer’s MAP; item loadings threshold >0.40. CFA (Sample 2): WLSMV estimator due to multivariate non-normality; fit assessed via RMSEA, SRMR, CFI, TLI; internal structure via AVE, CR, standardized loadings. Measurement invariance across gender via multi-group CFA (configural, metric, scalar; ΔCFI<0.01; scalar also ΔRMSEA<0.015). Reliability via Cronbach’s α and CR; test-retest via Pearson r over 2 weeks. Construct validity: one-way ANOVA for age/gender differences; partial correlations (controlling SES) between I/D subscales and external measures (parent- and teacher-reported) to evaluate convergent validity.

EFA (Sample 1): Data suitable (Bartlett χ²=686.78, df=45, p<.001; KMO=0.88). Parallel analysis supported two factors (λ1=5.18>0.78; λ2=0.74>0.37; λ3=0.24<0.27; λ4=0.03<0.16). MAP minimum at 2 factors (values: 1→0.06, 2→0.05, 3→0.06, 4→0.09, 5→0.14). I-type explained 32.60% variance; item loadings 0.69–0.87. D-type explained 28.64% variance; item loadings 0.56–0.86. CFA (Sample 2): Two-factor model fit well [χ²(34)=115.85, p<.001; RMSEA=0.08 (90% CI 0.06–0.10); SRMR=0.05; CFI=1.00; TLI=0.99], outperforming one-factor model [χ²(35)=455.99, p<.001; RMSEA=0.18 (90% CI 0.16–0.19); SRMR=0.10; CFI=0.98; TLI=0.97]. I-type AVE=0.64; loadings 0.76–0.84. D-type AVE=0.71; loadings 0.74–0.91. Interest–Deprivation latent correlation r=0.72. Measurement invariance (boys n=205 vs girls n=184): Configural supported (CFI=0.950; RMSEA=0.089). Metric invariance ΔCFI=0.002; RMSEA=0.082. Scalar invariance ΔCFI=0.000; ΔRMSEA=0.004. Reliability (Sample 2): Cronbach’s α total=0.90; I=0.85; D=0.88. Composite reliability (CR) total=0.95; I=0.90; D=0.92. Test-retest (2 weeks; n=82): total r=0.84; I r=0.76; D r=0.77 (all p<.001). Construct validity: No age or gender effects on I/D-YC scores. Partial correlations (controlling SES; Sample 3 unless noted): High-intensity pleasure: I r=0.16*; D r=0.14. Perceptual sensitivity: I r=0.23**; D r=0.11. Inhibition control (higher scores=greater deficits): I r=0.15*; D r=0.11. Approach/Withdrawal: I r=0.26***; D r=0.25***. Williams Creativity Scale: I r=0.45***; D r=0.50***. Supportive parenting: I r=0.32***; D r=0.34**. Shyness: I r=−0.22**; D r=−0.21**. Teacher-reported curiosity (Sample 4): I r=0.11; D r=0.06 (ns). Significance: *** p<.001; ** p<.01; * p<.05.

The study validates the Chinese I/D-YC as a reliable, two-dimensional measure of epistemic curiosity in preschool-aged children, addressing the need for culturally sensitive instruments in collectivist contexts. The robust EFA/CFA evidence and measurement invariance across gender confirm that the Interest and Deprivation constructs are distinct and generalize across boys and girls. Reliability and test-retest indices indicate stable measurement over time. Construct validity patterns suggest that I-type curiosity is more strongly associated with joyous exploration and sensitivity to novel stimuli (high-intensity pleasure and perceptual sensitivity), while both I and D types relate to approach behaviors, supportive parenting, and lower shyness, consistent with broader engagement in novel environments. Contrary to some adult findings, inhibitory control showed a modest positive association with I-type scores when measured as deficits (CHEXI), aligning with the notion that reduced inhibition may allow overt expression of curiosity within the Chinese cultural emphasis on harmony and self-restraint. Both I and D types correlated with creative inclination in preschoolers, suggesting that in early development, multiple pathways of curiosity may support creativity, potentially differentiating with age into distinct contributions to divergent thinking. Parent- and teacher-reported curiosity did not correlate, highlighting discrepancies in observational contexts and potential inadequacy of the older teacher scale to capture modern epistemic curiosity manifestations. Overall, findings support the instrument’s validity and underscore cultural and developmental nuances in how curiosity relates to temperament, self-regulation, parenting, and creativity.

The Chinese version of the I/D-YC scale replicates the original two-factor structure and demonstrates strong reliability, measurement invariance, and convergent validity for assessing epistemic curiosity in Chinese preschoolers. Results suggest that nurturing curiosity in early childhood, with less emphasis on strict inhibitory control and more supportive parenting, may foster well-being and creativity. The scale provides a valuable tool for future research on curiosity development and its cognitive and socio-emotional correlates in China. Future work should refine teacher-reported measures, incorporate objective behavioral tasks, and employ longitudinal designs to elucidate age-related differentiation of I- and D-type curiosity’s contributions to creativity and learning.

Teacher-reported curiosity (brief TQYC) did not align with parent-reported epistemic curiosity and may inadequately capture contemporary epistemic behaviors, indicating a need for improved teacher-focused instruments. The Approach/Withdrawal subscale showed modest reliability (α=0.63). All external measures were informant-reported; future studies should include objective behavioral paradigms and experimental tasks. The cross-sectional design limits inference on developmental trajectories; longitudinal research is needed to examine how I and D curiosity differentially relate to creative processes over time.